HCI Bibliography Home | HCI Conferences | HCII Archive | Detailed Records | RefWorks | EndNote | Hide Abstracts
HCII Tables of Contents: 89-1a89-1b89-2a89-2b91-1a91-1b91-2a91-2b93-1a93-1b93-1c

Proceedings of the Sixth International Conference on Human-Computer Interaction

Fullname:Proceedings of the Sixth International Conference on Human-Computer Interaction
Editors:Uuichiro Anzai; Katsuhiko Ogawa; Hirohiko Mori
Location:Tokyo, Japan
Dates:1995-Jul-09 to 1995-Jul-14
Publisher:Elsevier Science
Standard No:ISBN 0-444-81795-6 ISSN 0921-2647; hcibib: HCII95
Pages:1179+1067
Links:www.elsevier.com
  1. HCII 1995-07-09 Volume III. Analysis, Design and Evaluation in Human-Computer Interaction
    1. III.1 Interaction in Context Context in Interaction (Panel)
    2. III.2 Winning the Market of HMS (Panel)
    3. III.3 Interaction Design 1
    4. III.4 Interaction Design 2
    5. III.5 Interaction Design 3
    6. III.6 Screen Design 1
    7. III.7 Screen Design 2
    8. III.8 Screen Design 3
    9. III.9 Active Interface
    10. III.10 Evaluation and Analysis 1
    11. III.11 Evaluation and Analysis 2
    12. III.12 HCI Evaluation Methodologies
    13. III.13 Usability Engineering
    14. III.14 Cognitive Engineering
    15. III.15 Computer Modeling of Mental Processes
    16. III.16 Modeling 1
    17. III.17 Modeling 2
    18. III.18 Voices and Faces

HCII 1995-07-09 Volume III. Analysis, Design and Evaluation in Human-Computer Interaction

III.1 Interaction in Context Context in Interaction (Panel)

Interaction in Context -- Context in Interaction BIB 5-17
  William Edmondson; Jim Alty; Patrick Brezillon; Erik Hollnagel; John Meech; Katsuhiko Ogawa; Dan Suthers

III.2 Winning the Market of HMS (Panel)

Winning the Market of Human-Machine Systems (HMS) BIB 21-23
  Elena A. Averbukh; Toshio Fukada; Kensuke Kawai; Gunnar Johannsen; Rajko Milovanovich; Alberto Stefanini

III.3 Interaction Design 1

Different Approaches in Visual Interactive Software Construction BIBA 27-32
  Olivier Esteban; Stephane Chatty; Philippe Palanque
In this paper we focus on tools and environments for visual interface development that have been proposed recently. The aim is to introduce Whizz'Ed, an experimental editor for construction of highly interactive or animated applications in order to highlight the original contribution of the Whizz'Ed approach. Whizz'Ed provide an easy-to-use environment using elementary components for visual design and development of an highly interactive interface, allowing rapid constructions by non-programmer users.
Toward a Comprehensive Manipulation Solution on 3D Workspace BIB 33-38
  Nobuo Asahi; Kazuhisa Okada; Akira Maenaka; Eun-Seok Lee
Tailoring Non-Visual Interaction in a Graphical Environment BIBA 39-44
  C. Stephanidis; R. Gogoulou
The provision of alternative, non-visual interaction techniques is necessary in order to enhance non-visual interaction possibilities in a graphical environment, and fulfil the needs and preferences of blind users. This paper addresses issues related to the non-visual reproduction of a graphical environment and describes a dialogue configuration system (CONFIG) for tailoring non-visual interaction on the basis of individual blind user's needs and preferences, through 'easy-to-use' interactive facilities.
Command-Line Prediction and Error Correction Using Generalized Command-Line BIBA 45-50
  Yasuhiro Ueda; Toshikazu Nishimura; Michihiko Minoh; Katsuo Ikeda
A UNIX shell that maintains the history of not only command-lines but also "generalized command-lines" is proposed. A generalized command-line consists of a command name and argument symbols, and represents the syntax of a command. The system can detect an erroneous command-line and correct it, even if the system designer does not give such syntactic data to the system in advance. Moreover, by analyzing the history of generalized command-lines, the system can get information useful for command-line prediction and thus achieve better prediction.

III.4 Interaction Design 2

FRADS: A System for Facilitating Rapid Prototyping by End Users BIBA 53-58
  Irvin R. Katz
While user interface toolkits and managers facilitate prototyping by programmers, few systems allow nonprogrammers to create their own applications. In this paper, we report some techniques that bring prototyping to nonprogramming domain experts, namely professional test developers at Educational Testing Service. The Free-Response Authoring and Delivery System (FRADS) allows professional test developers to create dynamic, working prototypes of computer-based test questions. FRADS was designed to leverage nonprogrammers' experience with commercial graphics packages. Test developers create questions by importing graphics and other user-interface objects, choosing the tools to provide to students in responding to the question, and delineating -- via dialog boxes and specially designed graphical objects -- how the tools and provided interface objects interact. With FRADS, we explore how much "programming power" can be obtained by using direct, graphical specification of applications.
User Interface Development Environment for End Users: CUIDE BIBA 59-64
  Y. Tokuda; E. S. Lee; N. Shiratori
To cope with individual requirements of user interface (UI) from various classes of users, it is desirable to have UI development support environment which allows end users to develop UIs without designer's helps. The support of conventional methods and tools, however, focuses on expert of UI development. In this paper, we propose a UI development environment for end users, named CUIDE (Case-based UI Development Environment) in order to support UI development by end users, who have never developed any UI. CUIDE utilizes the case base of design and the case base of parts to make up for their lack of knowledge and experience of UI development.
Supporting Computer Users through Dialogue BIBA 65-70
  Akira Ito; Tadahiko Kumamoto; Tsuyoshi Ebina
The authors have been developing a dialogue-based user support system which assists a novice computer user in performing computer tasks. If a user gets into some trouble while using a computer, the system can provide help through natural language dialogue. The system understands a user's utterance and generates an appropriate response.
   The e-mail handling program XMH is used as target software. The system records each of XMH operations by the user, and maintains XMH information. To answer user questions, the system consults the current XMH state, and operational history of the user.
A Denotational Approach for Formal Specification of Human-Computer Dialogue BIBA 71-76
  K. Matsubayashi; Y. Tsujino; N. Tokura
This paper presents a formal approach to specify the human-computer interaction, especially the information exchanged between a computer and a user, using a notion derived from Denotational Semantics [8], which is originally developed for the formal description of the semantics of programming languages. As a first step, we applied the specification technique to existing application programs, and successfully obtained brief and concise specification of the meaning of user's input symbols to the system.
A Case-Based Method to Support Creative Design Incorporating Intention Recognition BIBA 77-82
  Takayuki Yamaoka; Shogo Nishida
In this paper, we describe a method and a system to support a creative design process. This method is characterized by its ability to infer the designer's intention based on case-based reasoning (CBR) methodology, so that the system can provide more useful and cooperative information on the current design.
Designing Interfaces for Computer-Based Assessments BIBA 83-88
  R. M. Kaplan; I. R. Katz
Over the last several years we have developed many new constructed-response items types. Part of our efforts in this process has been to develop, in conjunction with the item types, automatic or semi-automatic means for scoring these items. The development of scoring processes feeds directly back to the interface design process. Because creating complex intelligent applications to analyze a particular item can be time consuming, costly, and result in a process that cannot be generally applied, this approach to scoring these items is not a viable approach. An alternative to this is to constrain the item interface in such a way as to assist the scoring process while at the same time leaving the task realistic.
   We have shown two item types for which this type of development process took place. The first is a graphical task for architects. Rather than allow an architect to develop a completely open-ended solution, the item interface collects the same information while constraining the activities of the architect. Similarly, the second item, in the domain of writing, constrains the activities of the writer while collecting information like that which would be produced in a completely open-ended writing task.

III.5 Interaction Design 3

WMH Methodology for HCI Design BIBA 91-96
  Christian Coeffe
This paper aims at bringing up a practical methodology for Human Computer Interaction (HCI) design and a formalism for HCI description which have been successfully used in Air Traffic Management (ATM) field, namely for:
  • the French PHIDIAS programme developing the next generation of Air Traffic
       Controller Working Position to be installed in 1997 (Ref 1; Ref 2; Ref 3);
  • the European SWIFT project launched by the European Commission
       (DGVII/EURET1.5) (Ref 4; Ref 5). This so called WMH (What, Means & How) methodology is not tremendously innovative but it works and its usability by both operational and technical people has been proved. Due to a pressing need for better methods and tools in this part of system design, we think that what we see as a successful experience deserves to be made known to a wider circle of HCI design experts.
  • Guided Understanding for Problem Solving Process Using the Refining Self Explanation BIBA 97-102
      K. Kanenishi; Y. Yano
    In this paper, we presented the recognition model of self explanation. Self explanation consists of the reflection and the explanation production. By giving a self explanation, the understanding of the learner's problem-solving process progresses. We showed the necessity of the environment to make the self explanation lively. Then, according the way of supporting a self explanation, the learner who does agood to explanation is seem to understand an object well. We described construction of the self-explanation environment, 1) the explanation transformation module, 2) the evaluation module, 3) the planning explanation support module, 4) the advice control module 5) the interface module. And we described way to refine an explanation, 1) the insertion of the partial explanation, 2) the deletion of the partial explanation and 3) the division of the partial explanation.
       We proposed the graphic explanation environment. The graphic environment helps the understanding of the learner. Also, the environment of our system integrates the conventional ITS and the environmental CAI [7-9].
    A Strategy and Technology for Front End System Development BIBA 103-108
      Linda Candy; Ernest Edmonds; Susan Heggie; Bryan Murray; Nick Rousseau
    This paper describes an approach to the enhancement of existing software and the development of new applications based upon the premise that advanced software technology is not in itself sufficient to realise high quality usable systems. Development strategies designed to ensure quality must be accompanied by appropriate system architectures and effective implementation tools. We describe a technology and strategy that together enable the efficient development of user and task support systems in a wide variety of contexts.
    A Methodology for Developing New Interaction Techniques BIBA 109-114
      Deborah Hix; James N. Templeman; Ankush Gosain; Kapil Danderkar
    We present a methodology for inventing, implementing, and evaluating new interaction techniques. We illustrate use of this methodology using examples of some of the more interesting issues we encountered in developing a new interaction technique for head-coupled panning and zooming, called pre-screen projection.
    Basic Models for User Interface Design: Tasks, Users, Data, and Interaction Devices BIBA 115-120
      Chris Stary
    In this paper a representation scheme for the specification of the users of an interactive system and its functionality in a declarative as well as procedural way is introduced. According to this scheme a design methodology is given that does not end with the isolated treatment of interface components but rather provides a sound integration strategy.

    III.6 Screen Design 1

    The Effects of Realistic versus Unrealistic Desktop Interface Designs on Novice and Expert Users BIBA 123-128
      L. A. Miller; K. M. Stanney
    Metaphors are one tool which designers have used to bridge the gap between technology and the novice user. While metaphors work in theory, often the metaphor falls apart or breaks down during seemingly logical procedures. It would stand to reason that a metaphor which approximates the human-environment interaction, and remains consistent in the analogy, would be a more valid test of the benefits of a metaphor. In this study, novice and expert subjects were given a set of four computer-based, editing tasks, each of which was composed of six subtasks. The dependent variable was time to complete the task.
       It was found that novice subjects significantly benefited from the use of the realistic desktop design, while there were no significant differences in the performance of expert users.
    Rendering Calligraphy Words with 'Kasure' Variations BIB 129-134
      Qinglian Guo
    Dynamic Font: Its Concept and Generation Method BIBA 135-140
      K. Takayama; H. Kano; Y. Maeda; K. Misue; S. Hosogi; K. Sugiyama
    A new concept of "dynamic font" and its generation method [1] were proposed. The dynamic font was generated by intersecting some virtual writing implement with some virtual plane and moving the implement continuously in both space and time subject to the designed writing-motion.
       The writing-motion was defined by using the concept of "unit motions" [2] and this made possible a local and dynamic generation of motions and the fonts as if human wrote such fonts in real time [3].
       In order to build various motions and the fonts, a notion of operations on motions [2] was also contained. They included spatial operations such as scaling, translating, rotating, and tilting a motion. Structural operations of joining two motions and separating a motion into two enabled to generate any sequence of continuously connected cursive fonts.
       The writing-motion was represented formally as a sequence of the weighting coefficients for unit motions [2]. The sequence formed a "control polygon" geometrically and was used effectively to design the motion as well as the dynamic font.
       Several simulation examples were demonstrated by using an elliptic cone and a simple planar plane respectively as examples of the virtual writing implement and the plane for writing on. The figures showed fonts of really different typefaces were generated from similar control polygons except the rates and timings of putting the implements up and/or down and the shapes of the implements.
    A Spatial Data Structure for the 3D Graphical Facility Management System BIBA 141-146
      Takashi Tamada; Teruhiko Teraoka; Minoru Maruyama; Shogo Nishida
    Geographical information systems (GISs) are now used extensively in the analysis of environmental data, due to their capability to manage, manipulate, and display spatial data. However, most of the GIS applications (such as facility management system) possess data management structures that deal with only 2 dimensional space. The 3D system that enables users to explore the 3D space interactively and examine 3D spatial views of the environments would be very useful. In this paper, we propose a spatial data management method for a 3D graphical facility management system. Using this method, we have developed prototype 3D graphical management system that offers interactive operations with the 3D virtual city environments.
    Cryptographic Human Identification BIBAK 147-152
      Tsutomu Matsumoto
    Conventional password checking schemes have a disadvantage that an attacker who has correctly observed an input password can perfectly impersonate the corresponding user. To overcome it there have been proposed schemes such that a human prover knowing a secret key is asked a question by a machine verifier, who then checks if an answer from the prover matches the question with respect to the key. This paper presents practical interactive human identification schemes with novel ideas of displaying questions.
    Keywords: Authentication, Human-computer interaction, Passwords, Information security

    III.7 Screen Design 2

    Adjustment Mechanism for a Drawing System with Stationery Metaphors BIBA 155-160
      Naoki Kato; Natsuko Fukuda; Masaki Nakagawa
    This paper described an automatic adjustment mechanism for a drawing system employing a handwriting (pen) interface with stationery metaphors. With this mechanism the computer automatically carries out adjustments of the position or size of stationery metaphors at the stage of drawing. By including this adjustment mechanism in the drawing system it is possible to support the time-consuming task of making adjustments, and thus improve efficiency of the users.
    Analysis Tool for Skill Acquisition with Graphical User Interfaces Based on Operation Logging BIBA 161-166
      Nobuko Kishi
    Observing users working with a system is essential for better user interface design. One of popular observation methods is logging users' operations, i.e., having the system collect the users' operations automatically. However, it is known that operation logs, particularly the ones collected in the systems with graphical user interfaces (GUI), can be very large and difficult to extract meaningful information.
       We developed a set of tools, collectively named SimUI, to log and analyze users' operations on applications with GUI. SimUI's analysis tool uses a technique, named multi-step matching, to detect differences between two sets of operation logs. This paper describes the use of multi-step matching in a experiment, in which the subjects are asked to learn to use a bitmap editor. In the experiment, we observed that the 'matching rate,' measured by multi-step matching, can indicate various type of skill acquisition, compared to statistics about the task completion time and the frequency of command use.
    The Role of Screen Parameters in Visual Communication BIBA 167-172
      Masaaki Kurosu; Hitoshi Yamadera; Itaru Mimura
    One reason why the video conference system has not yet widely used is the lack of the evidence that will convince us its cost-performance compared to the face-to-face meeting. Reality is a key concept to describe the performance of the system. Previous researches on the video conference system had a tendency to focus on the hardware and software aspects of the system. But few of them studied the human side of the system, i.e. how the reality can be measured or how physical parameters may affect the degree of the reality. In this study, we tried to fix the independent and dependent variables relating to the reality and performed a psychological experiment on some of these variables.
    Re-Sampling of 3-D Object Range Data by Cube-Based Segmentation BIBA 173-178
      S. J. Wang; Y. Cai; M. Sato; H. Kawarada
    This paper proposes an efficient method for scattered range data re-sampling using uniform segmentation in 3-D space. The method combines the techniques of uniform cube-based segmentation, patch formation, and lattice points generation. After analyzing the scattered range data which formed by optical measuring machine, a preliminary processing is carried out by making the cross section in 3-D object space to form a group of cubes. Based on the cubes from preliminary processing, we generate lattice points in each cube. So using fewer re-sampling points, it is possible to reconstruct a smooth and complete 3-D object with parametric surface equations.
    Harmonic Curve Design BIBA 179-184
      John R. Rankin
    Many aesthetically appealing curves can be generated by using epicycloid and spirographic parametric curve equations. The difficulty in obtaining an interesting and appealing curve comes from having to choose good values for the coefficients in these equations. If a poor choice of coefficients is made the curve generated will have little regularity and a high amount of confusion. Since these curve types involve up to ten real coefficients the set of aesthetic curves is a set of measure zero on the infinite coefficient space. Rather than making random searches through coefficient space another approach which proved to be far more productive in discovering curves of aesthetic appeal was investigated. In this approach we looked at the Fourier Transform of the curves and performed editing on the coefficients in Fourier space. This approach is especially suited to the epicycloid and spirographic curve types which turn out to be simply low-order harmonic functions after suitable filtering and a rotation. As a result, curves can now be defined by free-form input and then cleaned up by this method to yield the nearest epicycle or spirograph.

    III.8 Screen Design 3

    GUIs and SUIs: More of the Same or Something Different? BIBA 187-192
      Alison Black; Jacob Buur
    Solid user interface (SUI) is a term coined in Japan to distinguish the user interface of products with embedded microprocessors like video cassette recorders, photocopiers and cellular phones from the Graphical User Interfaces (GUIs) of computer applications.
       SUIs are a neglected area in HCI research despite the substantial revenue earned from sales in both industrial and consumer applications. The great potential of SUIs that are easy to use is evident from the success of products such as the VideoPlus (VCRPlus) programmer, which has reduced the mystery of video programming in households world-wide.
       Despite the trend towards integrating products with SUIs and computer systems, we believe the distinction between GUIs and SUIs is worth preserving because user interaction with SUIs differs significantly from interaction with GUIs. And for many industrial and commercial applications SUIs offer distinct advantages over GUIs
       The objectives of this paper are to encourage increased attention to SUIs in the research community. We will identify characteristics that differentiate SUIs from GUIs and highlight the different methods for analysis, design and testing that SUIs require.
    Vision-Based Human Interface System with World-Fixed and Human-Centered Frames BIBA 193-198
      Kang-Hyun Jo; Yoshinori Kuno; Yoshiaki Shirai
    This paper presents a vision-based human interface system that enables a user to give 3D positioning commands by moving his hand. The user can choose either the world-fixed frame or the human-centered frame for the reference frame with which the system interprets his motion. If he uses the latter, he can move the object forward by moving his hand forward even if he changes his body position. The key for the realization of the system with this capability is vision algorithms based on the multiple view affine invariance theory. We demonstrate an experimental system as well as the vision algorithms. Appropriate uses for both frame cases are discussed.
    Fuzzy Reasoning Approach to Iconic Interface Design BIBAK 199-204
      Rungtai Lin
    Understanding how users recognize an icon is as complex as understanding visual perception itself Because the cognitive factors that affect the evaluation of icons have not been properly analyzed, the comprehension of pictorial symbols is typically ill-defined. The fuzzy approach seems to be a better way to provide a possible solution to this methodological problem. To remedy the disadvantage of the conventional approach, and to use the full potential of the computer as an aid to icon design, the purpose of this study is to propose a fuzzy graphic rating as a technique for evaluating the icon design. Although a fuzzy rating can be conducted using a pencil and paper technique, analysis is a difficult part through obtaining the rating in a pencil and paper method. In addition, iconic interface design is a highly graphic-oriented, so the graphic is playing an important role in design process. Therefore, based on the CAD technology and the fuzzy graphic rating as an approach to iconic interface design, this paper is intended to propose a more systematic treatment of iconic interface design than has hitherto been made. In this paper, a FUZRID (FUZzy Reasoning in Icon Design) system was implemented to provide designers with the predictive information using the fuzzy reasoning approach.
    Keywords: Fuzzy set theory, Icon design, Human factors
    Inferring Graphical Constraints from Users' Modification BIBA 205-210
      Takeharu Tanimura; Tsukasa Noma; Naoyuki Okada
    This paper presents a new approach to inferring graphical constraints in graphical editors. In our approach, users modify graphical objects interactively so that the objects have approximate geometric relations, and relevant graphical constraints are automatically inferred by comparing the states before and after the modification. This approach can grasp users' intention more accurately than existing approaches that infer constraints only from a single state of drawings.
    Enhancing Fold Manipulation Techniques BIBA 211-216
      Ying K. Leung; Richard J. King
    Folding is an effective means to overcome the inherent problem associated with displaying a large information space on a small computer screen. This technique has been successfully deployed in a variety of application domains, ranging from text editors and spreadsheets to information spaces with a tree structure hierarchy. However, this paper argues that the folding facilities currently available in many commercial products could be improved. It further proposes a number of features which may be used to enhance folding to enrich the human-computer interaction.
    Providing Diagram User Interfaces for Interactive Theorem Proving BIBA 217-222
      Jun Han; Tao Lin
    User friendly interfaces are essential for interactive theorem proving to be practical. We have argued that proofs can be naturally modelled as having graph structures. As such, diagrammatic presentation is seen as a natural medium for communicating proofs between the user and the theorem proving engine. In this paper, we present an approach to providing diagram user interfaces for interactive theorem proving.

    III.9 Active Interface

    Active Interfaces for Useful Software Tools BIBA 225-230
      Amedeo Cesta; Daniela D'Aloisi; Vittorio Giannini
    The majority of public domain software tools is endowed with programmer-oriented interfaces and, as a consequence, they are mainly devoted to "hackers" or specialized programmers. This paper concerns the development of an active interface used to make the utilization of a free software for filtering e-mail messages easy. A multi-agent implementation of an interface is proposed that from one side allows the user to easily specify his needs and from another actively reasons on the user's choices to improve the performance of the filtering process.
    Interacting with Real Objects: Real Object Interface and Transferred Object Interface BIBA 231-236
      Soichiro Iga; Michiaki Yasumura
    This paper describes the concept of "Real Object Interface (ROI)" and "Transferred Object Interface (TOI)". This idea is to create a virtual world that corresponds to a real world, and manage a computer by a real object by making one-to-one correspondence between the real object in the real world and the virtual object in the virtual world. Then we would test these concepts by making two application examples on a simple prototype system called "OuterDesk" and describe the result. Lastly, we discuss on its impact to new computer application areas.
    User Identification in Human Robot Interaction Using Identification Pendant BIBA 237-242
      Kaoru Hiramatsu; Yuichiro Anzai
    In this paper, we propose a user identification system in human-robot interaction, and design and implement the system. The identification system consists of transmitters called identification pendants and receivers equipped on personal robots. For signaling between pendants and receivers, we use infrared signals. With the system, a robot can identify users who are near the robot. Considering the issue of a user identification, we can design more convenient applications.
    Applying Personal Robots and Active Interface to Video Conference Systems BIBA 243-248
      Nobuyuki Yamasaki; Yuichiro Anzai
    Robots for personal use like current personal computers will appear in an office or at home in the near future. We call these robots personal robots. As of today, a personal robot can be thought as a small general-purpose autonomous mobile robot. For a personal robot interface, we have already proposed a new user interface concept: Active Interface. Active Interface does not only wait for users' explicit input but also tries to get information from users' implicit input and external environment. Based on the gathered information, it acts spontaneously and keeps the system in an advantageous condition for users. In this paper, we apply personal robots and Active Interface concept to a video conference system. The system should be made flexible and user-friendly. In brief, we design and implement a self-movable, flexible and user-friendly video conference system.
    An Object-Oriented GUI for the Specification of Robotic Systems BIBA 249-254
      Raymond K. Wong
    To facilitate the specification of robotic systems in a natural, incremental and systematic way, a graphical user interface (GUI) for the extended object-oriented model which supports the visual specification and modeling of the robotic systems as well as the computation of robot kinematics is described in this paper. This GUI provides an interactive environment from robot kinematics (low level) to robot activities (high level) in more user-friendly way when compared with most of the existing modeling tools.
    Augmented Interaction: Interacting with the Real World through a Computer BIBA 255-260
      Jun Rekimoto
    This paper discusses why traditional GUI is not adequate to support highly portable computers, and proposes a new HCI style called Augmented Interaction, which is concentraining on the user's real world activities. Situation awareness and implicit interaction are the two key ideas of this concept. We also report on the prototype system called NaviCam, which is based on the idea of Augmented Interaction.
    InfoBinder: A Pointing Device for a Virtual Desktop System BIBA 261-264
      Itiro Siio
    InfoBinder is a new wireless pointing device that provides an information-binding function in a virtual reality environment. Each device has a unique ID number and is mapped to an object such as a telephone directory in the computer system.

    III.10 Evaluation and Analysis 1

    An Object Oriented Methodology for Man-Machine Systems Analysis and Design BIBA 267-272
      A. Mahfoudhi; M. Abed; J-C. Angue
    Despite the recent progress in the domain of Man-Machine Interface engineering, several problems concerning the incompatibility between the information presentation to the user and his cognitive representation remain. This paper presents a new Task Object Oriented Description methodology (TOOD), especially adapted to the taking into account of the human factors for the specification of the Man-Machine Interfaces (MMI). A concrete application of this methodology was presented in the air traffic control context.
    An Analysis of Relationship between Human and Information System by Quantification Theory III BIBA 273-278
      Tsuneki Mukahi; Ken Murasugi; Tetsuo Ui
    The purpose of this study is to clarify the framework for understanding the relationship between human and advanced, large-scale information systems by analyzing the empirical data gathered using questionnaire. As the result, three axes of 'goodness of relationship between human and information system', 'attitude of organization toward the information system', 'stress felt by individuals in the information system' were found out. A new image of 'an advanced information-based organization' was obtained from these axes. It was also observed that such 'an advanced information-based organization' has been already realized in some organizations.
    Towards an Effective Subjective Measurement Method Based on Fuzzy Set Theory BIB 279-284
      Hiromi Terashita; Mieko Ohsuga; Futomi Shimono; Mamiko Toda
    The Design and Experiment of an Evaluation Function for User Interaction Cost in the Interactive Semantic Disambiguation BIBA 285-290
      Masaya Yamaguchi; Nobuo Inui; Yoshiyuki Kotani; Hirohiko Nisimura
    In this paper, we will design a function which evaluates the cost of user interaction that resolves the semantic ambiguity in natural language processing system. Through user interaction, the system shows some example sentences that express each meaning to let the user select one from them. We define the evaluation function based on semantic and syntactic features of verbs.
    An Analysis of the Human-Computer Interfaces to High-Energy Physics Control Systems at CERN BIBA 291-296
      J. F. Meech; P. Huuskonen; E. Wagner; M. Meri; J.-M. Le Goff
    This paper describes an investigation of the user-interfaces to high-energy physics experiments at CERN, Geneva. As part of this project a study of existing interfaces was carried out together with an analysis of operator characteristics and responsibilities. The results of this study are presented together with recommendations for the design of future user-interfaces for High-Energy Physics experiments.

    III.11 Evaluation and Analysis 2

    PDS Analysis for Evaluating Procedural Usability on Conversational Systems BIBA 299-304
      Akinori Komatsubara; Masayuki Kobayashi
    This study discusses whether we can evaluate procedural usability by checking only guidance. Four experimental systems were developed: The systems consist of same sequence but the guidance is designed differently. Human errors are compared among the systems. Based on the results of the experiments, the effects of guidance on procedural usability are to be discussed.
    Quantitative Evaluation of Media Quality by Method of Competitive Priority BIBAK 305-310
      Hiroshi Tamura; Jun Wu
    Method of competitive priority was proposed to evaluate quality of speech and images. Speech coded by ADPCM showed some priority to the one by PCM under competitive situations. The image quality was evaluated by presenting two speech words associated with talking head image of one of the two speeches. The effect of image presentation was compared with normal, ISDN (64 kbps) and the stop motion image. The effect was low with ISDN image. Some people were found insensible to ISDN or the stop motion image.
    Keywords: Speech image interaction, Image quality, Speech quality, Talking head video, Image compression
    Evaluation of Control Strategies in a Complex Space-Vehicle Control Task: Effects of Training Type BIBA 311-316
      Ravindra S. Goonetilleke; Colin G. Drury; Joseph Sharit
    The fundamental differences in operator control strategies in a complex task were evaluated in two training scenarios: in-the-loop training and out-of-the-loop training. Verbal protocols and performance measures revealed four types of complex control mechanisms dependent upon these two training approaches. The four types were display based control, open loop input control, closed loop input control, and an input-display control mix. Performance differences favored in-the-loop training, and led to the development of an open loop input control strategy. The overall results indicate that performance improvements may be achieved with operator training on the system dynamics and optimization aspects rather than operator training directed only at the optimization aspects. A "sitting by Nellie" approach such as watching an expert or watching an algorithm perform a task may be disastrous if the system dynamics are poorly understood. This study also suggests how operator strategies can be effectively used to design user-friendly aids which improve operator performance in complex control tasks.
    Development of the Analysis Support System for Incidents and Troubles; "ASSIST" BIBA 317-322
      Yuriko Yoshizawa; Keiko Mutoh
    For safe operations and works, it is important that human error not be made. So we strengthen equipment and educate workers for safety. It is necessary to provide a methodical education system and to take trouble data. So we need to make an integrated information system on human error and equipment troubles for supporting human error reduction activities in the field.
    Discount Video Analysis for Usability Engineering BIBA 323-328
      Mark H. Chignell; Tetsuro Motoyama; Venicio Melo
    Usability analysis has to be cost-effective. In this paper we discuss the role of video analysis in usability engineering, outlining an approach for making it easier and less time consuming. After reviewing the various approaches to usability analysis, the role of video in iterative design and evaluation is discussed. A method is proposed for simplifying video analysis by automating clip segmentation based on sound analysis and other techniques. This is followed by a brief description of a direct manipulation video editor that we are developing for usability engineers.
    User Interface Evaluation: Is It Ever Usable? BIBA 329-334
      Christelle Farenc; Philippe Palanque; Jean Vanderdonckt
    This paper has shown that the availability of guideline knowledge is not enough in order to provide helpful and efficient information to the agents involved in the design process of an interactive application. By showing the different points of view of those agents towards the UI we have explained why it is so difficult to provide relevant and easy to use information. Two different projects addressing this question are presented. The first one aims at evaluating and proposing solutions for the improvement of the interactive applications previously developed, while the second one aims at providing a set of tools for the automated design of interactive applications taking into account guidelines for user interface design. This paper has tried to give an answer at the question included in the title: without appropriate tools supporting both design and evaluation, it is no more allowed to follow the numerous guidelines currently available. Those tools are a necessary condition but of course not a sufficient one as the craft knowledge of user interface designers and ergonomists is far to be not worthy. User Interface evaluation: is it ever usable?

    III.12 HCI Evaluation Methodologies

    Software Tools for Evaluating the Usability of User Interfaces BIBA 337-342
      Sandrine Balbo
    In this article we propose a review of some different techniques and methods to evaluate the usability of user interfaces (UI). So far, the evaluation process has been mostly based on "craft techniques" [Long 89], but as we will demonstrate, formalisation is possible, and a few software packages in this domain will be presented.
       The models and techniques we will consider are:
  • general guidelines such as those proposed by [Smith 86, Nielsen 90, Bastien
       93],
  • the Cognitive Walkthrough [Lewis 90],
  • metrics [Whiteside 85, Bevan 94],
  • usability labs [Hammontree 92, Weiler 93],
  • predictive models [Young 90, Barnard 87],
  • automatic monitoring systems [Siochi 91, Balbo 94] and
  • critics [Lowgren 90, Kolsky 89]. We will present these methods and techniques around a taxonomy developed by Joelle Coutaz in [Coutaz 94], taxonomy designed to help in the choice of a method to evaluate UI. As well, we will highlight the role played by software tools for evaluating the usability of UI.
  • How Usable are Usability Principles, Criteria and Standards? BIBA 343-348
      J. M. C. Bastien; D. L. Scapin
    One evaluation method for human-computer interfaces, usually called expert-based evaluation, is defined as an informal method of usability analysis consisting of an analytic examination of a specified, prototyped or existing interface, with the goal of identifying ergonomic design flaws. It relies either on the evaluators' expertise (be they human factors specialists, system designers, software engineers, etc.) and/or on some human factors knowledge as available in documents such as general design guides [1], sets of guidelines [2, 3], checklists [4], standards (e.g., AFNOR, ISO, etc.), and heuristics [5, 6] or criteria [7, 8].
       All of these documents have been developed for the purpose of good human-computer interface design. Paradoxically, only a few of these documents have been evaluated in terms of their validity, thoroughness, reliability, effectiveness, and their ease of use by their potential users. The paper presents available data on these issues and research work focusing on the assessment of ergonomic criteria. The aforementioned issues are then discussed together with the research needed to develop a set of ergonomic criteria [8] into a full evaluation method.
    Usability is Quality of Use BIBA 349-354
      Nigel Bevan
    In a paper at HCI International 1991, Bevan et al (1991) asked "What is usability?", and distinguished between broad and narrow approaches to usability. This paper builds on that distinction, identifying the broad approach to usability with the higher level quality objective of "quality of use" (Bevan, 1995a). Quality of use should be the major design objective for an interactive product: does the product enable the intended users to achieve the intended tasks? This relates usability to business objectives and elevates usability from an optional extra to the prime design goal.
       The narrow approach is complementary and is concerned with the design of features of the product which are a pre-requisite for quality of use. The two different interpretations of usability lead to two approaches to the specification and evaluation of usability.
    Usability Evaluation: How Does It Relate to Software Engineering? BIBA 355-360
      Deborah Hix
    We present an integrated set of activities for development of both user interface and non-interface components in an interactive system. Within the context of these activities for both software and user interface engineering, we present several types of techniques for evaluation of usability. For each technique, we give examples, strengths and weaknesses, and results of its use at appropriate stages in the software engineering process.
    Standards and Software-Ergonomics Evaluation BIBA 361-366
      Harald Reiterer; Reinhard Oppermann
    There is an increasing need for practical and comprehensive evaluation methods and tools for conformance testing with standards. Practical means that the amount of time and resources must be manageable in software projects. Comprehensive means that the context of use has to be considered during the evaluation of user interfaces. The evaluation approach EVADIS III is such a practical and comprehensive one. In particular, it takes the context of use into consideration and provides computer support for the use of the evaluation procedure. It supports the evaluator during the evaluation process with detailed instructions. The first version of EVADIS III should be available at the end of 1995.
    Using Ergonomic Rules for Evaluation by Linguistic Ergonomic Criteria BIBA 367-372
      Francois Bodart; Jean Vanderdonckt
    Heuristic Evaluation is a well established method for evaluating a user interface (UI) of an interactive application [1]. This method belongs to the class of informal evaluation methods since the evaluation is performed on the basis of evaluator expertise and knowledge. One characteristic of heuristic evaluation is to guide evaluators by providing them a small set of predefined heuristics rather than a long list of general guidelines. Evaluators are given a widely applicable description of these heuristics so that they are general and largely applicable [2].
       In evaluation by ergonomic criteria [3], heuristics that have been judged too general are replaced by a taxonomy of eight main ergonomic criteria (i.e., guidance, workload, explicit control, adaptability, error management, consistency, significance of code, compatibility). These criteria could be decomposed into sub-criteria leading to eighteen elementary (sub-)criteria. These criteria present at least three advantages:
  • 1. they are precisely defined, especially when distinguishing between
        concurrent criteria;
  • 2. they have been experimentally tested and validated;
  • 3. they could be directly linked to useful ergonomic rules (or guidelines). In this paper, we introduce the evaluation by linguistic ergonomic criteria which is an extension of the latter method. First, the reasons that motivated this extension are presented by showing their definition. Second, first steps toward a complete evaluation method are outlined. Third, we exemplify this approach by detailing a UI evaluation report. Finally, we discuss our experience with this work in progress.
  • III.13 Usability Engineering

    A Teaching Method as an Alternative to the Concurrent Think-Aloud Method for Usability Testing BIBA 375-380
      P. R. Vora; M. G. Helander
    In this paper, we propose a teaching method as an alternative to the concurrent think-aloud (CTA) method for usability evaluation. In the teaching method, the test participant, after becoming familiar with the system, demonstrates it to a seemingly naive user (a confederate) and describes how to accomplish certain tasks. In a study that compared the teaching and the CTA methods for evaluating usability of human-computer interactive tasks, the results indicated that the number of verbalizations elicited using the teaching method far exceeded those elicited using the CTA method. Also, the concurrent verbalizations were dominated by the participants' interactive behavior and provided little insight into the participants' thought processes or search strategies, which were easily captured using the teaching method.
    Tools for Iterative User Interface Design: UI-Tester and OST BIBA 381-386
      Toshiyuki Asahi; Hidehiko Okada; Osamu Iseki; Ryoichi Matsuda
    A computer-aided iterative design environment is proposed. A usability testing tool "UI-tester" and a user interface design tool "OST" are integrated so that iterative design programs can be effectively implemented in the early stage of product development without usability expertise. A feasibility study being conducted on an ongoing facsimile product development project in this environment gives data showing that common erroneous pattern extraction functions of UI-tester are applicable for identifying user interface problems, and that user interface redesigns are completed without additional delays in the product development cycle.
    A Composite Measure of Usability for Human-Computer Interface Designs BIBA 387-392
      Kay Stanney; Mansooreh Mollaghasemi
    A methodology for formulating a composite measure of interface usability is provided. The measure integrates multiple usability criteria into a single measure by which designs can be directly compared. The primary advantages of the proposed approach are the ability to consider multiple criteria and to weight the importance of these criteria according to a particular company's priorities and requirements.
    Why Choose? A Process Approach to Usability Testing BIBA 393-398
      T. Kelley; L. Allender
    Two sets of prototype screens for a complex, computerized analysis tool were evaluated using a series of usability analysis techniques. The empirical, or experimental usability method identified more interface design problems of a severe nature than the other methods and gave a clear indication of which prototype design to choose for the final development process. While the individual walkthrough evaluation identified the most design problems overall, many of the problems tended to be of a less severe nature than were identified by the experimental method. The implications for selecting appropriate usability techniques and using them collectively, as a process, are discussed.
    Usability and Quality Control of Human-Machine Interaction BIBA 399-401
      Elena A. Averbukh
    High quality of Human-Computer Interaction (HCI) in terms of usability, users' acceptance and job satisfaction becomes more and more critical along with the increasing complexity of automation, requirements to its safety, quality, ecological friendliness and technology transfer (Johannsen, 1994, Shneiderman, 1992).
       This in its turn demands more sophisticated and systematic approaches to the quality control of HCI during all phases of systems' interfaces life cycle, i.e., from conceptual design to the development, evaluation, operation and possible redesign/reuse (Averbukh and Johannsen, 1994, Katai et al, 1991).
       This paper presents an integrated approach to the quality control of HCI and focuses mainly on the problems of so-called "in-process inspection" of the quality of human behaviour during interaction with the computer.
       From the integrative point of view both interface development tools and interface software systems themselves are persistent subjects of quality control and adaptation, as it is schematically shown in Fig.1.
       Modern Human-Machine Interfaces (HMI) are designed as distributed knowledge-based systems which contain
  • knowledge about the users, i.e. User Model,
  • knowledge about application domain, that is, e.g., in industrial control
       applications Technical System (TES) Model,
  • knowledge about their interaction (Averbukh et al, 1994). Different strategies for increasing the usability of interfaces by both on-line and off-line adaptation and further management of these knowledge structures which consider specific users' needs and expectations in concrete task situations (Averbukh, 1994).
       The functionalities embedded into the HMI which support this adaptation are also depicted in Fig.1.
       Several advanced architectural design paradigms for effective implementation of such "in-process inspection" functionalities are discussed. For this purpose, the appropriate criteria of their effectiveness are formulated.
       The multidimensionality and the dominant role of the User Modelling functionality in the frame of usability and quality control is analysed and discussed. Concrete options and interface quality control strategies based on User and Situation Modelling are specified.
       The application examples are given for supervisory control of technical systems, particularly for the chemical industry.
  • III.14 Cognitive Engineering

    Color Coordinate Supporting System with Navigating State of User's Mind BIBA 405-410
      Yasushi Yagi; Tomohiko Yagyu; Yoshihiko Hisamori; Masahiko Yachida
    In this paper, we propose a retrieval system for detecting color coordination that matches with a current state of user's mind. First, the user selects a few keywords such as wild and vivid, and the system extracts the initial candidates of color coordination from design database which correspond to keywords. Next, these initial candidates are narrowed down to candidates adapting to the state of user's mind. Remained candidates usually suit user's color preference. Thus it is not easy for the user to select one from these candidates. Therefore, finally, by changing the balance of color coordination, the user can find the favorite design of the color coordination with great satisfaction.
    Comparison between Three Human-Interfaces in Hospital Information System BIBA 411-416
      Kotaro Minato; Akira Endoh
    In this report, we describe a quantitative method to evaluate human-computer interaction (HCI) for the direct prescription order entry system in the hospitals. This method is based on a GOMS-like cognitive model for the interaction and the model is represented by a tree structure of five layers on goal-task hierarchies. Three different interfaces at university hospitals were compared by this method and the differences (similarity) among them were measured.
    Explaining Plant Design Knowledge through Means-End Modelling BIBA 417-422
      Pertti Huuskonen; Kari Kaarela
    We apply the multilevel means-end model of Rasmussen to explain the intentions behind design choices of industrial systems. We have extended a design tool, the Design++ environment, to capture higher level knowledge about the artefact under design and to explain this knowledge to the users. Means-end knowledge is encoded through relations added on top of the existing part-of and subclass hierarchies in the tool. Explanations are encoded as structured text in the objects and design rules. We have tested the explanations with a power plant application. Although still very primitive, this research prototype has confirmed that means-end information can be effectively used both to capture and explain design knowledge. We adopt a simplified view to design processes and decisions, placing our main emphasis in artefact modelling and explanation techniques.
    Method of Ecological Interface Design Applied to Interactive Diagnosis Support System BIBA 423-428
      Yoko Asano; Shun-ichi Yonemura; Hiroshi Hamada; Katsuhiko Ogawa
    This paper proposes a procedure for designing ecological interfaces suitable for interactive diagnosis support systems based on the perspectives of work analysis and interface design proposed by Rasmussen [1, 2]. Several of the perspectives proposed by Rasmussen are chosen and concrete transformation paths are created for them. The effectiveness of the proposed procedure is confirmed by designing and testing a support system for interactive diagnosis.

    III.15 Computer Modeling of Mental Processes

    Computer Analysis of Characteristics of Creative Thinking and Self-Esteem Level BIBA 431-435
      A. E. Kiv; V. A. Molyako; Stephen T. McHale; V. G. Orishchenko; I. A. Polozovskaya
    The present paper consists of a description of correlation between the level of creative abilities of a person and his self-esteem. The relationship is demonstrated between creative thinking level parameters and the self-esteem level. These dependencies may be used for the different mental groups comparison.
    Computer-Based Testing of Reflective Thinking: Executive Control of Erroneous Performance in 9 to 12 Year Old Children BIB 437-442
      Uri Shafrir
    The Creative Thinking Testing by Using of Testing Problems Based on Different Logical Schemes BIBA 443-447
      A. E. Kiv; V. A. Molyako; V. L. Maloryan; I. A. Polozovskaya; Zelina I. Iskanderova
    New methods of testing of creative thinking are considered. They based on a mathematical model of thinking processes described in [1, 2]. This study consists more general consideration of computer testing programs structure. We have shown that there is a possibility of measuring of creative thinking parameters by using of computer testing problems based on different logical schemes.
    From Novice to Expert Decision Behaviour: A Qualitative Modelling Approach with Petri Nets BIBA 449-454
      Matthias Rauterberg
    To support the human factors engineer in designing a good interactive system a method has been developed to analyze the empirical data of the interactive decision behaviour described in a finite discrete state space. The sequences of decisions and actions produced by users contain much information about the mental model of this user, the individual problem solution strategies for a given task and the underlying decision structure. We distinguish between (1) the logical structure, (2) the sequential goal structure, and (3) the temporal structure. The analysing tool AMME can handle the recorded decision and action sequences and come up automatically with an extracted net description of the task dependent decision model (the logical structure). This basis model was filled up with additional elements to reconstruct one empirical action sequence of an expert user. Four different models are presented and their predictive power discussed.
    Modeling and Simulation of Human Operator in Mental Task Handling Qualities BIBA 455-457
      Celestine A. Ntuen
    Effective human control of a complex system depends in part on the design of decision aids that have (high-level) models of human mental processes. Since the human operator utilizes the computer as a medium of interaction with the task environment, it is important that tightly coupled and interoperable human mental models be incorporated into HCI. This paper reports on a pilot study of simulating mental task performance in a desk top HCI platform.
    The Interface Improvement for the Creative Thinking Computer Testing BIBA 459-462
      V. V. Chislov; V. L. Maloryan; I. A. Polozovskaya; G. V. Shtakser; A. I. Uyemov; I. G. Zakharchenko; M. Athoussaki
    The interface design is one of significant components which the successful creation of computer testing programs depends from. This is especially important for children's creative thinking testing. A new approach to this problem that takes into account modern tendencies in interfaces design and the Language of Ternary Description [1] is supposed.
    Evaluating Human Operator Models in Tool-Based User Interface Design BIBA 463-468
      Maria Athousaki
    This paper describes the context of use of a software tool aiming to provide assistance for the ergonomic design of man-machine interfaces as well as to facilitate usability engineering. The tool is currently being developed by SIEM Ltd and constitutes one of the company's lines of activities towards innovative solutions in the area of user interface design and evaluation. The novelty of the tool being described is that it relies upon encoded knowledge of the human operator (user profile, mental model) and the tasks to be performed, to construct prototypical implementations of alternative potential designs. These, subsequently become subject to evaluation and usability testing.

    III.16 Modeling 1

    Associative User Modeling: A Neural Network Approach BIBA 471-476
      Qiyang Chen; A. F. Norcio
    This paper presents and discusses an approach of user modeling. A set of neural networks is utilized to store, maintain and infer users' task-related characteristics. Such networks function as associative memories that can capture the causal relationships among users' characteristics for the system adaptation. It is suggested that this approach can be expected to overcome some inherent problems of the conventional stereotyping approaches in terms of pattern recognition and classification. It can also avoid the complexity of truth maintenance in default reasoning that is required in previously known stereotyping approaches.
    Personality Engineering: Applying Human Personality Theory to the Design of Artificial Personalities BIBA 477-482
      Linda S. Endres
    To understand and explain our world, we often resort to "anthropomorphism," attributing human characteristics to things that are not human. For example, intelligence and friendliness are often perceived -- in varying degrees and by various definitions -- in computers and other machines. This perception of human traits in nonhumans can be interpreted as "artificial personality." Because anthropomorphic systems have proven controversial, more research is required to give us the insights needed to develop more compatible and effective artificial personalities. The purpose of this paper is to introduce the concept of "personality engineering" to enable us to better understand artificial personalities and ultimately to improve the interaction between human and artificial personalities. To accomplish this, personality psychology is recommended as a framework within which to design, develop, and evaluate any system or product. A brief example is given to illustrate how one human personality theory could be adapted for the study of artificial personalities.
    Using the Template Model to Analyse Interface Specifications BIBA 483-488
      C. R. Roast; J. I. Siddiqi
    This paper describes the application of a formal modelling technique within human computer interface design. The model described is termed the template model which is a system based model which explicitly identifies system abstractions (known as template abstractions) that have a defined relation to intended task and users' capabilities [4]. Using template abstractions is it possible to express and analyse general interface usability properties within a formal framework. We describe the application of the template model in the analysis of a wysiwyg-style word processor. In this way we are able to demonstrate the potential role of the model and the impact of formal analysis with a familiar yet non-trivial application. In the case of the word processor analysed, we are able to characterise system constraints that determine what user tasks are supported. In general, we argue that employing the template model helps explicate assumptions about interface requirements.
    Task Model -- System Model: Towards an Unifying Formalism BIBA 489-494
      Ph. A. Palanque; R. Bastide; V. Senges
    In this paper we have proposed a way for translating from a UAN description to a Petri net one. For every UAN construct, we have proposed a Petri nets which interpretation is the same as the one of the UAN description. Lastly, we have shown how those sub-Petri nets can be merged in order to model complex task models.
       This approach provides significant advantages:
  • the ambiguities of the task models are solved during the translation process.
       This process, while automated to a certain extend, is interactive as it asks
       the designers of the task models to make more precise their specification if
       needed. Afterwards, the task models can be mathematically validated by
       analysis the Petri net model;
  • as we use Petri nets for modeling the interactive application, cooperation
       between the model of the tasks and the model of the system can be
       mathematically checked, in order to ensure before implementation that the
       system will be able to perform all the user's requests included in the task
       models;
  • as the model of the system is embedded at run time, it can be used to provide
       contextual help about the behavior of the system, as described in [10].
       When the task model is also embedded within the same formalism, the help may
       be given not only with respect to the system behavior but also with respect
       to the user's task and goal.
  • III.17 Modeling 2

    Scenario Based Specification of Interaction Metaphors BIBA 497-502
      C. Stephanidis; C. Karagiannidis; A. Koumpis
    A scenario-based formal specification of interaction metaphors is proposed, which assists designers to develop interaction dialogues at a high-level of abstraction. A formalism has been developed which provides means for describing both application and metaphor semantics, as well as a descriptive model for the specification of interaction scenarios. The proposed approach has been applied for the construction of application scenarios in the domain of special education (i.e. supporting students with learning difficulties), so that different user interaction requirements and cognitive abilities are addressed.
    Cocktail-Party Effect with Computational Auditory Scene Analysis -- Preliminary Report -- BIBA 503-508
      Hiroshi G. Okuno; Tomohiro Nakatani; Takeshi Kawabata
    One of important and interesting phenomena in sophisticated human communications is the cocktail party effect: that even at a crowded party, one can attend one conversation and then switch to another one. To model it in a computer implementation, we need a mechanism for understanding general sounds, and Computational Auditory Scene Analysis (CASA) is a novel framework for manipulating sounds. We use it to model the cocktail party effect as follows -- sound streams are first extracted from a mixture of sounds, and then some sound stream is selected by focusing attention on it. Because sound stream segregation is an essential primary processing for the cocktail party effect, in this paper, we present a multi-agent approach for sound stream segregation. The resulting system can segregate a man's voice stream, a woman's voice stream, and a noise stream from a mixture of these sounds.
    The Effects of Rehearsal on Visual Memory BIBA 509-514
      Mamoru Umemura; Hiroshi Ichikawa; Kenichi Teguri
    Recently the development of information equipment including computers has advanced rapidly. Accompanying it, the opportunity of using the man-machine interface using CRT display has increased. It is considered that in the case of the works that are carried out by reading the information displayed on a CRT, the easiness of observing the displayed information exerts large influence to the accuracy and efficiency of the works, fatigue and others. In the works accompanied by danger, its mistaken cognition may sometimes be linked with serious accidents. Accordingly, the easiness of its cognition is an important subject in view of safety.
       The easiness of recognizing the information displayed on CRTs has been reported regarding the color, shape, size and so on of display [1]. Besides, investigation has been carried out on the information processing by humans after reading display, especially on the maintenance of short-term memory [2-3]. In order to take proper judgment and action after reading displayed information, it is necessary to make information easy to be maintained in short-term memory. As its means, there is rehearsal, and it has been said that it is effective for maintaining information. Also in real scene, in the case of the works accompanied by danger, the confirmation of information by voicing has been frequently carried out.
       Therefore, this study aims at quantitatively investigating the following items by experiment.
  • (1) The effect that rehearsal exerts to the time interval of presentation.
  • (2) The effect of the rehearsal carried but by voicing in the case of
        continuously presenting information on CRTs.
  • Mechanisms of Slips in Display-Based Human-Computer Interaction: A Model-Based Analysis BIB 515-520
      Muneo Kitajima; Peter G. Polson
    Computation Model for Human Communication BIBA 521-526
      Masahiro Hiji; Hiroshi Nunokawa; Masatoshi Miyazaki
    We have propose a computation model for human communication which has both dynamism and variety. And we propose a possibility to represent human communications as a figure of communication. This makes the computation model enable to model a dynamism in human communication smoothly. We design a programming language based on the computation model.

    III.18 Voices and Faces

    Delivering the Promise of Speech Interfaces BIBAK 529-533
      Charanjit K. Sidhu; Gerry Coyle
    We compiled a style guide that includes checklists and methods based on our experience of speech interface design. These have been applied in user trials and field evaluations throughout the development of Call Minder, a network-based telephone call answering service. Numerous improvements were made to the usability of the service as a result.
    Keywords: Speech interfaces, Product life cycle, Usability engineering
    VOICEDIC: A Practical Application of Speech Recognition Technology BIBA 535-540
      Kenji Kita; Kazuhiko Ashibe; Yoneo Yano; Hiroaki Ogata
    This paper presents a practical application of speech recognition technology, a voice-accessible or voice-activated dictionary, which is undertaken by the VOICEDIC project at our laboratory. The paper will show an outline of VOICEDIC as well as its advantages. We will also describe implementational issues required to attain high speech recognition performance.
    An Operation Analysis of an Address Input System with Speech Recognition BIBA 541-546
      Kazuhiro Arai; Osamu Yoshioka; Shigeki Sagayama; Noboru Sugamura
    Address input is one of the most important processes in business. This process is needed for managing customer information and the delivery of commodities. In order to reduce the time necessary for address input, we have developed an address input system that employs speech recognition. Users of this system can input addresses by using not only a keyboard and a mouse but also speech. The input time was measured for 25 adult novice users in order to test whether or not the speech recognition function is useful. The results revealed that speech recognition reduces the time required for address input. This paper describes the system outline and the experiment measuring input time.
    A Menu-Guided Spoken Dialog System and its Evaluation BIBA 547-552
      Mikio Yamamoto; Takashi Koike; Seiichi Nakagawa
    This paper describes a man-machine spoken dialog system that integrates speech recognition and menu-based natural language understanding technique. One of the difficulties in speech recognition is that the recognition rate is inversely proportional to the size of the set of acceptable sentences in the recognition system. That is, there is a serious tradeoff between the recognition precision and coverage. To resolve this problem, we employ a menu-based natural language understanding technique as the guide for spoken input. We developed the preliminary system and evaluated it. The experiment shows that the menu-based spoken dialog system is suitable for novice users.
    Face Observation Using an Active Camera BIBAK 553-558
      Qian Chen; Takeshi Fukumoto; Haiyuan Wu; Masahiko Yachida
    This paper describes a new approach of taking good images for the face recognition such as the extraction of facial expression. The face like parts in an input image are first extracted by comparing the skin color regions and the hair regions detected from the image with several pre-defined 2-dimensional face pattern models using the fuzzy pattern matching method. The 3 dimensional pose of the extracted face relative to the camera is estimated using the area and the center of the gravity of the skin color region and the ones of the hair region on the face, which is then used to guide an active camera changing its view point in order to taking the image where the face will appear in the desired pose.
    Keywords: Face recognition, Active vision, Face detection, Pattern matching, Pose estimation, Skin color
    Facial Features and Configurations Affecting Impressions of Faces BIBA 559-564
      Takashi Kato; Masaomi Oda; Misami K. Yamaguchi; Shigeru Akamatsu
    The present study investigated the relationships between physical characteristics and perceived impressions of human faces, namely fierceness and gentleness of the face. The study demonstrated that the spatial properties and configurations of facial parts that distinguish between fierce and gentle faces can be identified. Such information on physical characteristics of the face may be useful to interface designers who wish to incorporate faces into human interface in order to improve the intelligibility and/or perceived friendliness of human-computer communication.
    Anthropomorphic Media Approach to Human-Computer Interactive Communication Using Face Robot BIBA 565-570
      H. Kobayashi; F. Hara
    This paper deals with the realtime response face robot to demonstrate the integration of three functions of active human interface (AHI) that realizes interactive human-computer communication through an anthropomorphic modality of face robot. As the 1st function of AHI, the face robot recognizes the existence of a human being in her view by using his/her face image data and measures the distance between human being and herself, as the 2nd function of AHI, according to the distance obtained above, the face robot determines the facial expression to be displayed, and as the 3rd function of AHL the face robot realizes the facial expression on her face. The experiment accomplishes the realtime response of the face robot in terms of her facial expression to the distance recognized in 80ms per one recognition.